Microsoft Word - ETASR_V13_N4_pp11182-11190
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11182
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
Investigation of European Union Horizon 2020
Information and Communication Technology
Projects with the Social Network Analysis
Method
Deniz Sekerci
Department of Industrial Engineering, Yildiz Technical University, Türkiye
deniz.bukcez@gmail.com
Selcuk Alp
Department of Industrial Engineering, Yildiz Technical University, Türkiye
alp@yildiz.edu.tr (corresponding author)
Received: 17 April 2023 | Revised: 22 May 2023 | Accepted: 23 May 2023
Licensed under a CC-BY 4.0 license | Copyright (c) by the authors | DOI: https://doi.org/10.48084/etasr.5967
ABSTRACT
This study aims to examine the partnerships of the projects accepted in the Horizon 2020 grant support
program, which supports innovative Research and Development (R&D) projects of the European Union,
using social network analysis according to criteria such as participant countries, the project subjects in
which they participate, and the number of funds they receive. Social network analysis can be used in many
different areas, as well as in the analysis of collaborations and partnerships. Turkey's partnership has not
been examined in detail for such international multi-partner collaborations. While taking advantage of the
opportunity to follow the trend of technology by conducting innovative R&D studies with leading
organizations in many different strategic fields, the interest and participation rate of Turkey is increasing
day by day in such programs, offering opportunities to obtain grant incentives. Examining the network of
collaborations established so far in this field can provide various inferences and suggestions for strategic
partnerships.
Keywords-social network analysis; international collaboration; Horizon 2020
I. INTRODUCTION
All institutions that are aware of the impact of international
cooperation in the field of science and technology give priority
to strategic partnerships [1]. Social Network Analysis (SNA) is
used in many fields. Although there are studies that investigate
the projects and partnerships accepted in the EU Horizon 2020
program, which is the subject of this study, no studies on
Turkey's performance and cooperation have been found for this
program. This study aimed to fill this gap in the literature and
reveal what kind of cooperation should be prioritized for
Turkey to be more effective in such programs. To observe the
areas where countries are active, 1276 ICT projects are
classified into 20 different topics. The success of the countries
in coordinating calls was calculated using weighting analysis
and examined whether the calculated country achievement
scores related to the degree of centrality. Finally, the types of
institutions of central countries and Turkey were examined,
revealing the types of Turkish institutions that should be more
active in such programs. Studies have been carried out in many
different fields, such as education, health, software,
international and interinstitutional relations, etc., using SNA.
Authors in [2] aimed to understand how the accuracy of
centralization behaves according to various amounts of error in
the dataset, investigating a large number of sample networks by
adding a controlled amount of error. The results showed that
accuracy decreases as the amount of error increases. In [3], a
study was conducted in the software industry using SNA,
focusing on topics such as monitoring architectural changes,
examining the effects of elements, and developing automated
tools for architectural analysis. In [4], SNA was applied to
obtain design principles to form a clinical team, based on the
interactions that are important to increase information flow and
achieve the intended results. In [5], SNA was used to explore
patterns of collaboration and seek advice between schools,
showing that while school networks are interconnected, overall
interschool collaboration and advice-seeking activities are at
low levels. In [6], a study was conducted on Clean
Development Mechanism (CDM) projects, suggesting that the
status of a country on the network is a sign of its strength in the
entire network. The results showed that the participating
organizations could decide which countries are more attractive
to invest in the CDM market. In [8], successful projects and
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11183
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
their partnerships in EUREKA were investigated using SNA,
aiming to map successful projects of Central and Southeast
European countries and identify the best-performing countries
between 2002 and 2009, publishing a report on the results of
participation in the 7th Framework Program (FP7). According
to the findings, 86% of the collaborations in the 6th Framework
Program (FP6) were not renewed in FP7, and universities and
research organizations "provided an advantage because they
were centrally located in the joint participation network of the
framework programs" [9]. In [10], SNA was used to examine
the international collaborations of the countries participating in
the Horizon 2020 program, showing that countries with a high
degree of centrality obtained large funding. This study also
showed that high centrality (degree, closeness, and
betweenness) is not an indicator of success rate. In [11], the
capabilities of the Italy/Calabria region were mapped with
SNA, demonstrating that its capacity in issues such as climate
action, food and agriculture, and information and
communication technologies is strong.
II. SOCIAL NETWORK ANALYSIS
A social network is a social structure consisting of
individuals or organizations connected by one or more specific
types of dependency, such as friendship, kinship, mutual
interest, financial exchange, etc. [12]. SNA aims to explore
social structures using networks and graph theory. Connected
structures are characterized by nodes, which are individual
actors, people, or things within the network. In addition, the
ties that bind them are characterized by edges or links
(relationships or interactions). The ability of SNA to be used in
a wide variety of domains is due to the flexibility of the nodes
and ties. Although primarily involving relations between
people, ties between groups or organizations and relations
between nation-states or international alliances can also be
examined with SNA [7]. The purpose of SNA is not only to
map and measure the relationships between nodes, but also to
understand the structure of a network, draw conclusions about
the impact of the relationship on an actor, and understand the
defining characteristics, structures, and consequences of the
relationships between them [13]. Examples of social structures
commonly visualized with SNA are social media networks,
information circulation networks, friendship networks, business
networks, working relationships, collaborations, kinship, and
disease transmission. In these networks, nodes are usually
visualized as points, and ties are visualized as lines [14].
A. Social Network Analysis Metrics
Using many different metrics in the SNA, inferences can be
made about the location and status of each individual in the
network, the status of the most important actors, and the
general characteristic structure of the network. In this context,
the centrality criterion and its breakdown within itself are
evaluated as node-based criteria in the graph structure. Apart
from this, criteria such as cluster coefficient, cliques, diameter,
density, etc., provide information about the general
characteristic features of the graph. The centrality metric seeks
answers to the questions of which actor is the most important
on the graph and is the most common and effective metric to
determine its importance. The centrality metric is expressed in
three dimensions: the ability of an actor to communicate with
others, his ability to control others, and his closeness to others
[15]. According to these approaches, the centrality metric is
examined under different sub-metrics, each showing different
information about the node. This section examines the degree
centrality, the closeness centrality, the betweenness centrality,
and the eigenvector centrality. The number of links a node has
with other actors in the network means the number of lines
related to what is called Degree Centrality [8]. Actors with a
high degree centrality are the most visible actors and have more
direct contact with other actors on the network [16]. Although
degree centrality is expressed as the number of edges
connected to the node in undirected networks, it is handled as
internal and external degrees in directed networks. The number
of incoming connections to the relevant node represents the
internal degree, the number of outgoing connections represents
the external degree, and the node degree is the sum of these
two values. In social networks, this metric is based on the
principle that the actor with the most connections in real life is
the most important. In other words, if the node has a higher
degree centrality, it has more importance in the network. In
directional networks, internal degree centrality represents the
popularity of an individual in the social network, and external
degree centrality represents the individual's sociability [17].
Closeness centrality refers to how close an actor is to the
other actors in the network. The main idea is that an actor's
ability to interact quickly with others depends on its central
location [16]. This metric can be used to identify the actors in
the best position that can affect the entire network most
quickly. When calculating the closeness centrality of a node,
the average of the shortest path lengths of that node to all other
nodes in the network is calculated, and the inverse of the
obtained value is taken. Betweenness centrality indicates which
nodes are more likely to be in communication paths between
other nodes. In other words, betweenness centrality is not
concerned with the proximity of a node to other nodes, but with
its location on the shortest path between other nodes [17].
When calculating betweenness centrality, first it should be
found how many shortest paths exist between a pair of nodes
(x, y), then count how many of the shortest paths between these
two nodes contain the v node, and a ratio is obtained by
dividing these two numbers. After this process is calculated for
all pairs of nodes, the sum of the calculated ratios is found as
the betweenness centrality of node v.
Eigenvector centrality is related to a node's connections
with other nodes of high importance. A node with many
connections does not need to have high eigenvector centrality,
and a node with high eigenvector centrality does not
necessarily need to have many connections. This approach is
related to the ability to make many friends in high places in
real-life networks [17]. A node's eigenvector centrality is
affected by the importance and weight of other nodes to which
it is connected. Eigenvector centrality depends on both the
quality and the number of connections. If a node has few but
high-quality connections, its eigenvector centrality may be
greater than a node with many but average-quality connections
[18]. To define the diameter of a graph, it is necessary to first
define the shortest path, which is the minimum number of links
required to connect two nodes in the network. The diameter of
a graph is defined as the longest of the shortest paths between
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11184
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
nodes in the network and is one of the key SNA metrics that
gives information on the size of the network in general [17].
Density is obtained by dividing the total number of
connections an actor has by the total number of possible
connections that he can have. A high result indicates that the
network is dense. Dense networks are likely to be found in
small and stable communities with few outside contacts and a
high degree of social interaction. On the contrary, loose social
networks tend to thrive in larger, unstable communities that
have many external contacts and exhibit a relative lack of
social cohesion [19]. Density values can take values between
[0, 1]. Density will approach 1 if there is a strong relationship
between the nodes of the graph and 0 if there is a weak
relationship. It is a useful metric used to make comparisons and
inferences. The clustering coefficient is a measure of the
probability that two partners of a node will be partners. The
basis of this metric is the concept of transitivity. It is
considered that there is a high probability of a relationship
between the x and z nodes in a graph if there is a relationship
between the x and y and a relationship between the y and z
nodes. A complete graph is formed when all nodes are related
to each other [18].
III. HORIZON 2020
Horizon 2020, is the European Union's research and
innovation grant support program that operated between 2014
and 2020. The main purpose of the program, with a total
budget of 80 billion euros, was to strengthen Europe in all areas
by encouraging technology development and multi-
international cooperation. This program aimed to enable
Europe to produce world-class science, eliminate barriers to
innovation, and facilitate the public and private sectors to work
together to deliver innovation [22]. Different types of legal
entities participated in Horizon 2020, such as universities,
Small and Medium Enterprises (SMEs), large companies,
research institutions, and Non-Governmental Organizations
(NGOs). The biannual work programs at Horizon 2020
included funding opportunities and calls. The minimum
number of participants and the requirements for the projects
varied according to the type of project. The main condition
sought in all applications to the Framework Programs, except
for individual projects, is cooperation. In general, at least three
independent organizations from at least three different EU
member states or associated countries should take part as
partners in Horizon 2020 projects.
Turkey participated in this program since the EU 6th
Framework Program. The EU 6th Framework Program took
place between 2002 and 2006, was replaced by the EU 7th
Framework Program covering the years 2007-2013, continued
from 2014 to 2020 as the Horizon 2020 program, and was
followed by the Horizon Europe program that covers the years
2021-2027. Turkey's representative TÜBİTAK transferred 451
million euros to this program, as each country participates by
giving a contribution. Therefore, organizations in Turkey were
allowed to apply for this fund program. Each country aims to
receive at least the amount of contribution it gives as a result of
the international cooperation it establishes [24]. Today,
technology development has proven to be very important in the
analysis and processing of the volume of scientific information
produced [25]. Horizon 2020 aimed to drive Europe's economic
growth by combining research with innovation, focusing on
three key pillars of scientific excellence, industrial leadership
and competitiveness, and societal challenges. This study
focused on the field of ICT, located in the facilitator and
industrial technologies section, under industrial leadership [23].
European Commission appoints independent experts who
evaluate project proposals and project operations and monitor
program and policy processes [24]. Projects with evaluation
scores above a threshold are ranked. Projects with higher scores
are entitled to be funded if they do not exceed the budget
allocated for the applied call. The Community Research and
Development Information Service (CORDIS) [27] is the main
source of results for projects funded by the European
Commission's framework programs for research and
innovation. CORDIS has a rich and structured public repository
containing all project information from the European
Commission, such as project fact sheets, participants, reports,
outputs, and links to open-access publications. CORDIS has
published a dataset on the European Union Open Data portal,
including information on projects and participants funded by
the EU under the Horizon 2020 framework.
IV. METHODS AND APPLICATION
A. Data and Visualization
This study used data from CORDIS. Datasets are updated
monthly on the CORDIS portal, and all projects accepted in the
Horizon 2020 program are publicly published. This study
analyzed data from 1276 projects with a total of 11384
partnerships. The dataset for each participant is listed with its
CORDIS registration number (RCN), project ID, project
abbreviation, the role of the organization in the project
(participant or coordinator), the name of the organization, the
type of organization, the grant amount given by the European
Commission (in EUR), and the country in which the
organization is located. Figure 1 shows a network where the
countries with the highest degree-centricity (with the most
diverse partnerships in all ICT projects) are positioned to be the
most centrally located and have the largest symbols.
Fig. 1. Network image of Horizon 2020 ICT projects created by 92
countries.
This study selected a total of 31 countries, including EU
countries, England, Switzerland, Norway, and Turkey, which
are the main participants of the Horizon 2020 program, to make
a more specific analysis. In this context, when the network
analysis graph was drawn again, a clearer visual was obtained,
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11185
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
as shown in Figure 2. The countries in the most central
positions are Italy, Germany, England, Spain, Greece, and
France. This image also highlights the strength of the bonds,
showing the countries that have the strongest ties to the central
actors, which are Belgium, Switzerland, the Netherlands, and
Austria.
Fig. 2. Network image created with 31 countries in Horizon2020 ICT
projects.
B. General Features of the Network
Table I shows the number of nodes, the number of links, the
density degree, the average distance, and the diameter values of
the undirected network consisting of 31 countries.
TABLE I. NETWORK FEATURES
Feature Value
Nodes 31
Ties 908
Density 0.976
Average distance 1.024
Diameter 2
There are 31 actors and 908 ties in the network. The density
of the mesh was found to be 0.976. Since this value is very
close to 1, it can be said that this network has a high density
and the actors are connected by strong ties. Diameter, another
basic metric that gives information about the size of the mesh,
took the value 2. So, the longest of the shortest paths in the
node is at a distance of 2. Furthermore, the average distance
between two actors was 1.024. As a result, it can be thought
that the network is not very large, but it can be said that the
density of the network is very high and therefore the actors are
positioned very close to each other.
C. Analysis of Social Network Analysis
The positions of the countries were examined by
performing SNA centrality metrics and clustering coefficient.
Centrality values were calculated to examine the positions of
the actors in the network. Table II shows the values of the
countries in degree, betweenness, closeness, and eigenvector
centrality, and their rankings according to these values. In
Figure 3, the trend lines increase linearly according to degree
centrality. Since the ranking of degree centrality and closeness
centrality are the same, the two lines appear as a single line of
green color. Around this linear line, the line highlighted in red
indicates the betweenness centrality, and the line highlighted in
dark blue indicates the eigenvector centrality ranking.
TABLE II. RANKING OF COUNTRIES BY DEGREE OF
CENTRALITY
Degrees Betweeness Closeness Eigenvector
IT 72 IE 366.295 IT 0.0091 IT 0.181
DE 70 IT 284.502 DE 0.0089 UK 0.181
UK 69 DE 271.051 UK 0.0088 ES 0.179
ES 67 UK 223.356 ES 0.0087 EL 0.179
FR 66 FR 193.535 FR 0.0086 DE 0.175
EL 66 ES 187.545 EL 0.0086 FR 0.175
PT 62 AT 175.547 PT 0.0083 PT 0.171
IE 60 EL 160.905 IE 0.0082 BE 0.167
BE 56 PT 157.637 BE 0.0079 NO 0.163
CH 55 CH 94.77 CH 0.0079 PL 0.163
NL 55 NL 78.423 NL 0.0079 CH 0.162
NO 53 BE 77.43 NO 0.0078 NL 0.159
AT 51 NO 67.887 AT 0.0076 AT 0.158
PL 51 LU 61.306 PL 0.0076 FI 0.157
FI 49 FI 36.588 LU 0.0075 IE 0.156
LU 49 PL 30.391 FI 0.0075 HU 0.155
BG 46 BG 28.532 BG 0.0074 RO 0.154
CZ 42 CY 12.027 CZ 0.0071 CZ 0.147
SE 41 RO 11.719 DK 0.0071 SI 0.144
DK 41 SE 11.126 SE 0.0071 HR 0.142
SI 40 SK 10.411 SK 0.0070 SE 0.14
SK 40 CZ 9.471 SI 0.0070 SK 0.138
HR 39 SI 4.273 HR 0.0070 DK 0.136
TR 35 HR 3.82 TR 0.0068 TR 0.122
LT 30 LT 0 LT 0.0066 LT 0.117
MT 23 MT 0 MT 0.0063 MT 0.092
Fig. 3. Representation of country rank change by centrality metrics.
As can be seen in Table IV and Figure 3, when looking at
the country rankings according to their degree centrality, Italy
developed the most cooperation by establishing connections
with 72 out of 92 partners in ICT projects, while Germany,
England, Spain, and Greece followed. On the other hand,
Turkey established partnerships with 35 different countries and
ranked 29th.
Ireland, Italy, Germany, England, and France are the
countries with the highest values of betweeness centrality.
Ireland ranked 8th in degree centrality and 1st in betweenness
centrality. This shows that Ireland is an effective actor with the
ability to control the flow of information in the network. In
addition, other countries that stand out in the chart and jumped
into betweenness centrality are Austria, Sweden, and Turkey.
Turkey has risen to the 20th place in betweenness centrality.
This shows that Turkey's ability to cooperate with a wide range
of actors is less than its capacity to influence the flow of
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11186
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
information in the network by taking the shortest paths. As
stated earlier, closeness centrality is related to the sum of the
shortest path lengths of an actor to other actors in the network.
Turkey was ranked 29th. When the eigenvector centrality is
examined, Italy, England, Spain, Greece, and Germany are
seen as the actors with the higher values. This metric, since the
importance of the actors with whom the actors are connected is
included, shows that these countries have a high degree of
relationship with the important actors. Although Norway and
Poland have lower rankings in other metrics, they are among
the top 10 countries in the eigenvector centrality ranking. This
shows that Norway and Poland are partnering with strong
partners, even if they do not establish more partnership
relations than the more centralized countries in the network. In
addition to that, Ireland and Finland showed a dramatic
decrease in this ranking, which was one of the most striking
changes in the chart. Turkey is seen to be in the 29th place in
this ranking. To examine the positions of the actors in the
network, the centrality values were calculated first. Table III
shows the values of the countries in degree centrality and their
ranking.
TABLE III. RANKING OF COUNTRIES BY DEGREE
CENTRALITY
Country
Clustering
coefficient
nPairs
Number of
neighborhoods
DE 40.28 435 30
ES 42.06 435 30
IT 42.86 435 30
FR 43.00 435 30
UK 43.24 435 30
EL 44.89 435 30
NL 46.19 435 30
BE 46.29 435 30
CH 47.42 435 30
PL 47.89 435 30
IE 48.26 435 30
FI 48.27 435 30
PT 48.41 435 30
AT 49.19 435 30
DK 49.66 435 30
EE 50.03 435 30
CY 50.08 435 30
LU 50.15 435 30
BG 50.36 435 30
RO 50.70 435 30
HU 50.71 435 30
SE 51.96 406 29
NO 53.29 406 29
SI 53.33 406 29
CZ 53.47 406 29
SK 54.39 406 29
LV 54.41 406 29
LT 57.85 378 28
HR 57.87 378 28
TR 61.70 351 27
MT 85.77 210 21
The second column, "nPairs", represents the number of all
possible pairs of actors in the neighborhood formed by the
neighbors with whom the actor is in contact or, in other words,
all possible connections. Neighbor numbers are given in the
fourth column. When the clustering coefficient of each actor is
ranked from the smallest to the largest, it is seen that
Germany's neighbors are connected with the least intensity,
followed by Spain, Italy, France, England, and Greece. In
addition, Malta, Turkey, Croatia, and Lithuania were last on the
list, and their neighbors also bonded highly. The fact that the
actors with low clustering density are those with a high degree
of centrality indicates that these actors are the most active in
the network and their clustering densities are low because their
neighbors are not as active as themselves. On average, actors
with higher rankings have been reported to have lower
clustering coefficients on average [26].
D. Partnerships of Turkey
Table IV shows the countries with which Turkey
cooperates the most. Turkey has realized the higher
cooperation with countries with the highest degree centrality.
This may be a natural consequence of those countries'
maximum number of partnerships in this program.
TABLE IV. TURKEY’S TOP TEN PARTNERSHIPS
Country
Number of
partners
Country
Number of
partners
ES 99 EL 38
DE 67 PT 27
IT 54 NL 24
FR 47 BE 23
UK 43 AT 21
Increasing the number of collaborations with Ireland, which
has the highest degree of betweenness centrality and, as a
result, the ability to affect the information flow at the highest
level, can enable Turkey to achieve greater success in this field.
In addition, increasing cooperation with countries with high
eigenvector degrees can enable Turkey to meet successful
partners in important positions. Norway and Poland, which are
not included in Table IV but have achieved a slight jump in
eigenvector centrality, are actors that can be considered in this
regard. Furthermore, as a result of the examination made using
clique analysis, Switzerland was found to be the ninth country
participating in the clustering and showed closeness to the most
central partners. Considering the current situation of Turkey,
increasing its ties with Switzerland may allow it to be more
included and positively affect its performance.
E. Partnerships of Turkey
The acceptance rates of all ICT calls were calculated by
dividing the number of projects accepted within its scope by
the total number of applications, and then inversely, the
importance of that call was obtained. The coordinator partner is
responsible for establishing and coordinating the project
consortium. Additionally, the coordinator has the largest role in
writing a good project proposal. For this reason, during the
calculation of the success rates of countries by weight analysis,
the number of times the countries participated as coordinators
was considered.
ICT calls are classified under 20 different headings
according to subject areas and the calculated country success
scores were distributed according to the classification made
under these headings. In the weighting analysis management,
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11187
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
the Acceptance Rate of ICT-x Call (CAR) was calculated first
by:
��� =
����
����
(1)
where NNAx is the number of applications accepted in the
opened ICT-x call, TNAx is the total number of applications in
the opened ICT-x call. The k-th country's coordinator
achievement score is given by:
�
� = ∑
���
�
���
��
��� � = 1,2, … ,31 (2)
where NCCkx is the number of coordinator ships from the k-th
country on the x-th call. As can be seen in Table V, the top 5
most successful countries according to the total success score
are Spain, England, Germany, Italy, and France, while Turkey
ranks 24th in this list. This situation shows that Turkey does
not take an active role as a coordinator in the program.
Although Germany and Spain seem to be the partners with the
highest scores on an equal number of topics, Spain's high
success, especially in the "Horizontal ICT innovation action",
made it the first in the ranking of the most successful countries
in total scoring. When examining the countries with which
Turkey has cooperated the most, it is seen that there is a strong
amount of cooperation with the first six most successful
countries according to the weight analysis. However, Finland,
the seventh most successful country, is not among the top ten
countries with which Turkey has developed a great deal of
cooperation. For this reason, further cooperation between
Finland and Turkey may be an important step in obtaining
successful project applications.
TABLE V. ACHIEVEMENT SCORES ACCORDING TO THE WEIGHTING ANALYSIS OF COUNTRIES
5
G
A
n
e
w
g
e
n
e
r
a
ti
o
n
o
f
c
o
m
p
o
n
e
n
ts
a
n
d
s
y
st
e
m
s
A
d
v
a
n
c
e
d
c
o
m
p
u
ti
n
g
a
n
d
c
lo
u
d
c
o
m
p
u
ti
n
g
A
I
a
n
d
d
ig
it
is
in
g
E
u
r
o
p
e
a
n
in
d
u
st
r
y
a
n
d
e
c
o
n
o
m
y
C
o
n
te
n
t
te
c
h
n
o
lo
g
ie
s
a
n
d
in
fo
r
m
a
ti
o
n
m
a
n
a
g
e
m
e
n
t
C
r
o
ss
-c
u
tt
in
g
a
c
ti
v
it
ie
s
C
y
b
e
r
se
c
u
r
it
y
E
u
r
o
p
e
a
n
d
a
ta
i
n
fr
a
st
r
u
c
tu
r
e
:
H
P
C
,
b
ig
d
a
ta
a
n
d
c
lo
u
d
te
c
h
n
o
lo
g
ie
s
H
o
r
iz
o
n
ta
l
IC
T
i
n
n
o
v
a
ti
o
n
a
c
ti
o
n
IC
T
c
r
o
ss
-c
u
tt
in
g
a
c
ti
v
it
ie
s
IC
T
k
e
y
e
n
a
b
li
n
g
t
e
c
h
n
o
lo
g
ie
s
In
n
o
v
a
ti
o
n
a
n
d
e
n
tr
e
p
r
e
n
e
u
r
sh
ip
s
u
p
p
o
r
t
In
te
r
n
a
ti
o
n
a
l
c
o
o
p
e
r
a
ti
o
n
a
c
ti
o
n
s
M
ic
r
o
-
a
n
d
n
a
n
o
-
e
le
c
tr
o
n
ic
te
c
h
n
o
lo
g
ie
s,
p
h
o
to
n
ic
s
N
e
x
t
g
e
n
e
r
a
ti
o
n
i
n
te
r
n
e
t
(N
G
I)
P
la
tf
o
r
m
s
a
n
d
p
il
o
ts
R
e
sp
o
n
si
b
il
it
y
a
n
d
c
r
e
a
ti
v
it
y
R
o
b
o
ti
c
s
a
n
d
a
u
to
n
o
m
o
u
s
sy
st
e
m
s
S
u
p
p
o
r
t
a
c
ti
o
n
s
S
u
p
p
o
r
t
to
h
u
b
s
T
o
ta
l
su
c
c
e
ss
ES 33 14 36 87 335 0 23 65 1193 36 18 15 6 43 264 28 2 87 0 10 2295
UK 0 21 18 20 261 8 0 9 805 6 11 0 9 12 190 0 5 80 14 0 1469
DE 40 124 37 n 202 8 21 51 305 44 16 10 9 63 170 12 0 82 0 8 1202
IT 22 42 27 63 136 20 31 37 404 42 22 7 30 18 184 12 0 148 0 5 1250
FR 8 100 31 108 49 8 12 5 540 6 25 12 3 39 79 6 7 65 0 9 1112
EL 23 23 22 79 189 13 13 93 128 21 14 7 11 22 188 0 0 23 0 2 871
FI 8 4 0 29 45 4 0 0 306 15 12 3 0 7 35 5 0 19 0 6 498
AT 0 36 0 31 60 13 3 25 169 12 20 0 3 11 71 0 7 31 0 0 492
NL 3 10 0 40 102 0 6 9 141 6 24 0 11 30 41 0 0 40 0 10 473
BE 3 32 22 34 74 8 4 0 23 12 14 0 0 32 88 0 7 20 0 2 375
IE 0 21 5 9 73 0 0 19 99 0 4 0 21 0 86 7 2 0 9 0 355
PT 3 4 0 14 8 13 15 13 152 0 0 0 9 0 74 8 0 0 0 2 315
SE 3 10 0 35 26 0 0 0 148 0 0 0 9 3 25 0 0 48 0 0 307
DK 0 10 0 5 8 0 8 0 163 0 0 0 0 6 25 0 5 27 9 2 268
NO 5 10 9 13 26 0 0 13 130 24 7 0 9 0 11 0 5 0 0 3 265
PL 0 0 5 0 20 8 0 0 178 0 0 0 0 0 8 0 0 6 0 3 228
HU 0 0 0 0 0 0 8 0 153 0 0 7 0 0 15 0 0 0 0 0 183
EE 3 10 0 0 0 0 0 0 143 0 0 7 0 0 3 0 0 0 0 0 166
CH 0 15 0 8 14 12 0 35 5 0 0 0 7 0 38 0 0 9 0 3 146
SI 0 0 0 9 0 0 0 9 99 0 0 0 0 0 12 0 0 11 0 0 140
CZ 0 0 0 0 8 0 0 5 41 0 0 0 0 0 5 0 0 0 0 0 59
LU 0 0 0 0 0 0 0 0 41 0 0 0 0 2 8 0 0 0 0 0 51
TR 0 0 0 0 0 0 0 0 41 0 0 0 0 0 0 0 0 0 0 0 41
CY 0 0 0 0 0 0 0 0 23 0 4 0 0 0 0 0 0 9 0 0 36
LT 0 0 0 7 0 0 0 0 18 0 0 7 0 0 0 0 0 0 0 2 34
SK 0 0 0 0 0 0 0 0 23 0 0 0 0 0 8 0 0 0 0 0 31
RS 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 15
RO 0 0 0 0 5 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 14
BG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 0 0 0 0 0 5
F. Comparison of Country Achievement Scores with SNA
Metrics and Country Success Percentage
The relationship between country achievement scores,
degree centrality, and country success percentages was
examined. The success percentage of the countries, which is
the ratio obtained from the total number of project applications
made in the Horizon 2020 ICT field and the number of
accepted project applications, was taken from the EU Horizon
2020 indicator panel [26]. Table VI shows the degree
centrality, success percentages, and success scores of the
countries.
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11188
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
TABLE VI. DEGREE CENTRALITIES, SUCCESS RATES, AND
SUCCESS SCORES PER COUNTRY
Country Degree centrality Success rates Success scores
CH 55 17.90% 144
CZ 42 17.42% 59
BE 56 16.17% 374
AT 51 15.65% 492
FR 66 15.38% 1114
NO 53 15.25% 264
FI 49 15.21% 51
NL 55 15.21% 474
DK 41 14.81% 308
DE 70 14.68% 1280
IE 60 14.65% 354
LT 30 14.07% 35
LU 49 13.61% 498
EL 66 13.43% 873
SE 41 13.16% 268
UK 69 12.86% 1469
PL 51 12.73% 226
SI 40 12.32% 31
CY 44 12.06% 36
ES 67 12.02% 2293
PT 62 11.92% 316
SK 40 11.68% 140
IT 72 11.51% 1251
HU 46 10.93% 183
EE 45 10.66% 167
RO 45 10.40% 14
TR 35 7.85% 41
BG 46 6.61% 5
CH 55 17.90% 144
CZ 42 17.42% 59
BE 56 16.17% 374
Correlation analysis provides information on the direction
and power of the relationship between two variables. A
correlation analysis was conducted to examine the relationship
between degree centrality, success percentages, and
achievement scores of the countries. To understand whether
there is a relationship between the variables, the following
hypotheses were established:
H0: There is no relationship between the two variables
H1: There is a relationship between two variables
To determine the result of the hypothesis test, the
correlation Table and the P-value were examined, and Figure 4
shows the results.
Fig. 4. Country’s degree centrality, success rates, and success scores
In the correlation analysis, if the P-value is less than 0.05,
the H0 hypothesis is rejected and the H1 hypothesis is accepted.
As can be seen in the figure above, there is a relationship
between degree centrality and country success scores.
However, the country's success percentage is not related to its
achievement score and degree centrality.
Italy, Germany, England, Spain, and France were the top
five countries in terms of degree centrality and country success
scores, while Switzerland, Czech Republic, Belgium, Austria,
and France are the most successful countries in country success
percentages. The reason the country's success percentages have
such a different ranking is that these countries have a good
acceptance rate even if they are not as active as the most
centralized countries in the ICT program. France, which is
among the top five countries in both comparisons, has achieved
both high cooperation and a good coordinator experience,
although it has made far fewer applications than other central
countries.
G. Comparison of The Types of Institutions Receiving the
Funds by Country
Table VIII compares the types of the most successful
organizations in Germany, England, France, Spain, and Italy,
which are the most central countries, and Turkey.
TABLE VII. COMPARISON OF THE MOST SUCCESSFUL
ORGANIZATION PERCENTAGE
Country University
Private
sector
Research
institute
Public
institution
Turkey 10% 80% 10% 0%
Italy 40% 30% 40% 0%
Germany 30% 50% 20% 0%
England 50% 30% 10% 10%
Spain 40% 20% 40% 0%
France 20% 50% 30% 0%
As can be seen in Table VII, research institutes and
universities, as well as the private sector, are actively involved
among the institutions that receive the highest funding in the
most central countries. This situation can offer important clues
about the issues Turkey needs to improve itself. Although only
10% universities and 10% research institutions are on the list of
the most successful in Turkey, most of them are private sector
institutions. In the most centralized countries, the sum of
universities and research institutions makes up half and even
more than half of the list. Universities and research institutes,
which mostly carry out studies at the stage of basic research
and the creation of new knowledge in a technology field, take
an important role in the first steps of new knowledge and
discovery and become an actor that directs the course of
technology.
V. CONCLUSION AND RECOMMENDATIONS
SNA can be used in many different sectors and fields and is
considered one of the powerful methods in examining
collaboration and partnership relations. Conducting a network
analysis provides a holistic view of the position of the partners
in the network and information about the relations between
them. This study used SNA to examine the projects that were
accepted in the field of ICT in the Horizon 2020 program. The
results showed that Italy, Germany, England, Spain, and France
were the countries with the highest degree and closeness
centralities. Ireland ranked first in betweenness centrality. The
fact that Ireland achieved a high value in betweenness
centrality shows that the country is an important factor that can
control the flow of information. However, Ireland is not on
Turkey's top partnership country list. Therefore, increasing
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11189
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
cooperation with Ireland may improve Turkey's position. In
addition, new collaborations can be developed with other
potential partners in important places by establishing
cooperation with Norway and Poland, which have rising
rankings in eigenvector centrality. In addition, clustering
coefficient and clique analysis were performed, and the
percentages of actors' neighbors being neighbors to each other
and which actors were closer to each other were examined. As
a result, it was seen that the clustering coefficients of the
countries could be low if they have high degree centrality. This
situation is related to the fact that the neighbors of the central
countries are not as active as themselves and do not establish
much connection. The clique analysis showed that the central
countries, with Spain and Germany in the first place, achieved
the first merger by showing closeness. Switzerland, which is
not on Turkey's top ten partnership countries list, was seen to
establish close relations with the central countries by being
included in the grouping in the 9th place. Turkey was found in
the 25th place and was characterized as an isolated actor.
Turkey can consider Switzerland as another country that should
increase its cooperation. Another country that stands out is
Sweden. Although Sweden has a low rank in degree centrality,
its rise of ten places in the clique analysis shows that it also has
strong relations with central European countries. These results
show, with exceptions, that geographical closeness also affects
collaboration.
In Horizon 2020 projects, it is an undeniable fact that the
coordinator actor plays an essential role in the success of the
project. The coordinator success of the countries was calculated
under 20 different headings by weighting analysis. Spain,
England, Germany, Italy, and France are the most successful
countries, respectively. Whereas Turkey is seen to have
developed high partnerships with the first six countries in the
coordination success ranking, more cooperation with Finland,
the seventh most successful country, can lead to successful
projects. Degree centrality, country success scores, and country
success percentages were compared, showing that there is only
a relationship between degree centrality and country
achievement scores. France is the only country in first place in
all three rankings, has established partnerships with a high
number of different countries, has developed coordinator skills,
and is almost as successful as the other central countries, even
though it makes much fewer applications.
Finally, the most centralized countries and Turkey's most-
funded institution types were compared, showing that
universities and research institutions in other countries have
much more active participation than universities and research
institutions in Turkey in Horizon 2020 ICT projects and have
an important role in the success of the country. Participants in
Turkey are mostly private sector organizations. More active
participation of universities and research institutions, which
carry out basic research and play an important role in
introducing new knowledge and technology, will increase
Turkey's success in the program and will be an important step
in becoming a technology developer country. In the future, this
study can be diversified with different and more in-depth
statistical analyses by focusing on different social network
analysis metrics.
REFERENCES
[1] M. Laaziri, S. Khoulji, K. Benmoussa, and K. M. Larbi, "Information
System for the Governance of University Cooperation," Engineering,
Technology & Applied Science Research, vol. 8, no. 5, pp. 3355–3359,
Oct. 2018, https://doi.org/10.48084/etasr.2156.
[2] S. P. Borgatti, K. M. Carley, and D. Krackhardt, "On the robustness of
centrality measures under conditions of imperfect data," Social
Networks, vol. 28, no. 2, pp. 124–136, May 2006, https://doi.org/
10.1016/j.socnet.2005.05.001.
[3] N. Vrček, I. Švogor, and P. Vondra, "Social network analysis and
software evolution: A perspective method for software architecture
analysis," in Proceedings of the ITI 2013 35th International Conference
on Information Technology Interfaces, Cavtat, Croatia, Jun. 2013, pp.
309–314, https://doi.org/10.2498/iti.2013.0543.
[4] D. Meltzer et al., "Exploring the use of social network methods in
designing healthcare quality improvement teams," Social Science &
Medicine, vol. 71, no. 6, pp. 1119–1130, Sep. 2010, https://doi.org/
10.1016/j.socscimed.2010.05.012.
[5] C. Sinnema, A. J. Daly, Y.-H. Liou, and J. Rodway, "Exploring the
communities of learning policy in New Zealand using social network
analysis: A case study of leadership, expertise, and networks,"
International Journal of Educational Research, vol. 99, Jan. 2020, Art.
no. 101492, https://doi.org/10.1016/j.ijer.2019.10.002.
[6] M. J. Kang and J. Park, "Analysis of the partnership network in the clean
development mechanism," Energy Policy, vol. 52, pp. 543–553, Jan.
2013, https://doi.org/10.1016/j.enpol.2012.10.005.
[7] L. C. Freeman, The development of social network analysis: a study in
the sociology of science. Vancouver, Canada: Empirical Press, 2004.
[8] B. Divjak, P. Peharda, and N. Begičević, "Social network analysis of
Eureka project partnership in Central and South-Eastern European
regions," Journal of Information and Organizational Sciences, vol. 34,
no. 2, pp. 163–173, Dec. 2010.
[9] Directorate-General for Research and Innovation (European
Commission), Fraunhofer ISI, Oxford Research, and Science Metrix,
Study on network analysis of the 7th Framework Programme
participation: methodological annex. Luxembourg: Publications Office
of the European Union, 2015.
[10] A. Bralić, "Social Network Analysis of Country Participation in Horizon
2020 Programme," in Proceedings of the Central European Conference
on Information and Intelligent Systems, Varaždin, Croatia, Sep. 2017.
[11] A. Morisson, C. Bevilacqua, and M. Doussineau, "Smart Specialisation
Strategy (S3) and Social Network Analysis (SNA): Mapping
Capabilities in Calabria," in New Metropolitan Perspectives, 2020,
https://doi.org/10.1007/978-3-030-52869-0_1.
[12] W. Stanley and F. Katherine, Social Network Analysis: Methods and
Applications. New York, NY, USA: Cambridge University Press, 2009.
[13] D. Z. Grunspan, B. L. Wiggins, and S. M. Goodreau, "Understanding
Classrooms through Social Network Analysis: A Primer for Social
Network Analysis in Education Research," CBE—Life Sciences
Education, vol. 13, no. 2, pp. 167–178, Jun. 2014, https://doi.org/
10.1187/cbe.13-08-0162.
[14] A. J. Masys, Ed., Networks and Network Analysis for Defence and
Security. Cham, Switzerland: Springer International Publishing, 2014.
[15] J. Scott, Social Networks: Critical Concepts in Sociology. London, UK:
Routledge, 2002.
[16] W. Stanley and F. Katherine, Social Network Analysis: Methods and
Applications. New York, NY, USA: Cambridge University Press, 2009.
[17] Z. A. Rachman, W. Maharani, and Adiwijaya, "The analysis and
implementation of degree centrality in weighted graph in Social
Network Analysis," in 2013 International Conference of Information
and Communication Technology (ICoICT), Bandung, Indonesia, Mar.
2013, pp. 72–76, https://doi.org/10.1109/ICoICT.2013.6574552.
[18] Q. Zhang, "The Changes of China Earthquake Disaster Emergency
Network Based on SNA," The Frontiers of Society, Science and
Technology, vol. 2, no. 7, Aug. 2020, https://doi.org/10.25236/FSST.
2020.020720.
Engineering, Technology & Applied Science Research Vol. 13, No. 4, 2023, 11182-11190 11190
www.etasr.com Sekerci & Alp: Investigation of European Union Horizon 2020 Information and Communication …
[19] P. Trudgill, Investigations in Sociohistorical Linguistics: Stories of
Colonisation and Contact. Cambridge, UK: Cambridge University Press,
2010.
[20] R. A. Hanneman and M. Riddle, Introduction to Social Network
Methods: Table of Contents. Riverside, CA, USA: University of
California, Riverside, 2005.
[21] F. Grando, D. Noble, and L. C. Lamb, "An Analysis of Centrality
Measures for Complex and Social Networks," in 2016 IEEE Global
Communications Conference (GLOBECOM), Washington, DC, USA,
Sep. 2016, pp. 1–6, https://doi.org/10.1109/GLOCOM.2016.7841580.
[22] Directorate-General for Research and Innovation (European
Commission), Horizon 2020 in brief :the EU framework programme for
research & innovation. Luxembourg: Publications Office of the
European Union, 2014.
[23] M. Hemmati and H. Hosseini, "Effect of IT Application on Project
Performance Focusing on the Mediating Role of Organizational
Innovation, Knowledge Management and Organizational Capabilities,"
Engineering, Technology & Applied Science Research, vol. 6, no. 6, pp.
1221–1226, Dec. 2016, https://doi.org/10.48084/etasr.769.
[24] "Horizon Europe," Anasayfa | Ufuk Avrupa. https://ufukavrupa.org.tr/en.
[25] K. Benmoussa, S. Khoulji, M. Laaziri, and K. M. Larbi, "Web
Information System for the Governance of University Research,"
Engineering, Technology & Applied Science Research, vol. 8, no. 4, pp.
3287–3293, Aug. 2018, https://doi.org/10.48084/etasr.2154.
[26] S. Borgatti, M. Everett, and L. Freeman, "UCINET 6 for Windows:
Software for social network analysis," Analytic Technologies, Harvard,
MA, USA, Jan. 2002.
[27] "European Commission Webgate," R&I Organisation Profiles - Qlik
Sense. https://webgate.ec.europa.eu/dashboard/sense/app/a22d6695-
65d1-4f7a-a06f-b5bf3f3cc59c/overview.